Evolving Threat Detector

Identifying changes in resistome relative to closest relatives and contextualising results

To run/test create a conda environment (or virtualenv) and run:

pip install -r requirements.txt --editable .

Then you should be able to run:

python etd.py

Make sure you add any new requirements you've added to the requirements.txt file

If you write then tests can be run as follows:

pytest

Alternatively to run the full set of tests including different python versions, pyflake code style, and building the documentation run:

tox

Overview

Output Structure

The main output directory is either specified by the user or defaults to the input genome name followed by the UNIX timestamp

test_genome_1576349112
├── mash
│   └── mash_distances.tsv
├── rgi
│   ├── test_genome.json
│   └── test_genome.txt
├── related_isolate_geospatial_analysis_of_relatives
└── unique_to_isolate
    ├── amr_gene1
    │   ├── phylogenetic
    │   └── genomic_context
    └── amr_gene2
        ├── phylogenetic
        └── genomic_context

External Dependencies

• MASH
• PPLACER
• HHMER
• e-utils

Database Preparation

The database can be built from CARD prevalence sequences i.e. a directory containing a set of directories for each taxa you want to include.

card_prevalence
├── genomes                                 # directory containing all CARD-prevalence genomes
|   |── Klebsiella_pneumoniae_NCBI_May2020
|   |   ├── NZ_NGWN01_wgs.fa
|   |   └── NZ_NIDM01000036.1_plasmid.fa
│   └── Klebsiella_oxytoca_NCBI_May2020
|       ├── NZ_CP029128.1_chromosome.fa
|       └── NZ_CP033845.1_plasmid.fa
├── rgi_results                             # directory containing all RGI outputs with the same name as the genomes
|   |── Klebsiella_pneumoniae_NCBI_May2020
|   |   ├── NZ_NGWN01_wgs.txt
|   |   └── NZ_NIDM01000036.1_plasmid.txt
│   └── Klebsiella_oxytoca_NCBI_May2020
|       ├── NZ_CP029128.1_chromosome.txt
|       └── NZ_CP033845.1_plasmid.txt
├── card                                    # directory containing the version of CARD canonical and prevalence used
|   |── card-data
|   └── card-prevalence
└── etd_db                                  # etd specific generated files
    |── etd_db_index.json                   # index for database
    |── card_prev.msh                       # mash sketch of all the genomes in CARD prevalence (i.e. genomes folder)
    |── genome_trees                        # directory containing all generated genome phylogenies
    |   ├── genome_trees_index.json         # index linking accessions to their specific tree
    |   ├── Klebsiella_pneumoniae_NCBI_May2020.mashtree
    |   └── Klebsiella_oxytoca_NCBI_May2020.mashtree
    └── amr_phylogenies                     # directory containing all clustered CARD+CARD-Prevalence phylogenies
        ├── amr_phylogenies_index.json      # index linking AROs to clusters
        └── cluster1.tree
        └── cluster2.tree

Detailed Workflow

Todo

Extract context using faidx (12.7kb average length of transposon): 2021.01.10.426126 Visualisation of neighbourhood with https://github.com/gamcil/clinker

Analysis of context: https://github.com/wtmatlock/flanker